European AI cloud providers Nebius, Nscale and Verda are building infrastructure with NVIDIA Vera CPUs, NVIDIA said in an October 20, 2026, update. The announcement details CPU-focused work across cloud services, server hardware and software platforms as AI agents create demand for processing beyond the GPUs that run models.
Key Takeaways
- Nebius, Nscale and Verda are building with Vera, while five server manufacturers are developing Vera-based systems.
- Data-processing companies reported strong early test results, but the update provided no numerical performance benchmarks.
- Agent-platform activity spans planned adoption, infrastructure under development and an existing use case at Nous Research.
Table of Contents
- NVIDIA Vera CPUs target the work around AI models
- Server options expand as data platforms test workloads
- Agent platforms show different stages of adoption
- Conclusion
NVIDIA Vera CPUs target the work around AI models
NVIDIA framed Vera’s role around the tasks agents perform after receiving a request: retrieving information, executing code, using software tools and evaluating results. Those operations depend on CPU processing alongside the GPUs responsible for running models.
The company says Vera is designed to maintain responsiveness while handling thousands of tasks simultaneously. It positions the processor as a way for providers to expand services and improve use of data-center capacity, although the update does not quantify those benefits.
The European infrastructure discussion also names Aleria as part of the broader sovereign AI ecosystem supporting customers that want control over their data and operations.
Server options expand as data platforms test workloads
Bull, Dell Technologies, HPE, Lenovo and Supermicro are building Vera-based systems. NVIDIA says the processor can be deployed in standalone CPU servers or paired with its GPUs in accelerated systems, offering different configurations for high-performance computing and AI data centers.
ClickHouse, Cloudera, Duck Labs, Elastic, Polars, Redpanda, SAP and Starburst observed strong results in early testing across analytics, extract-transform-and-load processing and semantic search, according to NVIDIA. The company attributed the gains to per-core performance, LPDDR5X memory bandwidth and its Scalable Coherency Fabric, without supplying numerical comparisons.
Agent platforms show different stages of adoption
The software examples range from evaluation to active use. Atlassian is looking to Vera to speed the processing used to build semantic search indexes for its AI products. Inherent has tested Vera and is considering adoption for environments where agents assist engineers with novel scientific research.
Nous Research, by contrast, already uses Vera to authenticate agent identities and validate actions before execution, NVIDIA said.
Baseten-Blaxel, Daytona, E2B, Modal, Prime Intellect and Vercel are building on Vera for cloud-based agent sandboxes. These isolated environments let agents execute code, run tools and test results. NVIDIA argues that startup speed becomes particularly important when sandboxes are provisioned millions of times a day.
Conclusion
The update broadens Vera’s announced role across European cloud infrastructure and agent software. Its examples remain at different stages, from early tests and intended adoption to systems under development and existing operational use.
Read More News
Disclaimer
This FinanceVelix article provides general information, not personalized financial, investment, legal, or tax advice. Market prices, interest rates and economic conditions can change; investments carry a risk of loss. Facts reflect the material available when published and may change. Check important information independently and consult a qualified professional when needed. Mentioning an asset, company, or service is not an endorsement.
